Most owners think turnover is a hiring problem. Sometimes it is. More often, the decision to leave gets made in the first few weeks, long before anyone hands in a resignation letter.
You've seen the people who have already made that decision. It's 4 o'clock and their desk is spotless. Everything is packed. They're sitting there waiting for the clock. Ask what they're up to and you get, "Nothing, just waiting."
They stay because they have to. The people you want stay because they want to. The difference usually gets decided in onboarding.
Why the first weeks decide so much
Think about what a new hire is really doing in the first few weeks. They're learning the job, sure. They're also quietly answering one question: did I make the right choice?
Every day gives them evidence. Does the training make sense to them? Does their manager seem to know them? Does the work fit how they like to work, or are they forcing themselves through it? Someone who has to adapt far from their natural style all day gets mentally exhausted and drained, often by mid-afternoon. Do that for a few weeks and the answer to the question becomes obvious.
And when someone leaves in their first 90 days, they take a lot with them: the time it took to recruit them, the time it took to train them, and the hours their manager spent on both. The rest of the team covers the gap until you start over.
The call center that blamed its applicants
A large call center came to us with turnover over seventy percent. Most people quit inside ninety days, which meant they left before they finished onboarding for the more complex accounts. The company almost never saw a return on anyone it hired.
They had already diagnosed the problem. They were hiring bad people, and they wanted help hiring better ones.
We didn't accept the diagnosis. We went looking for the root cause.
Same interview, same classroom, different people
This call center handled customer service for several very different companies. Some needed people who could work through complicated technical problems with patience and precision. Others needed warm, engaging people who built a relationship on every call. Two completely different kinds of work. Two completely different kinds of person.
Every applicant was interviewed exactly the same way. And the interviewers were highly extroverted. Without realizing it, they rewarded candidates who interviewed the way they did: high energy, quick rapport, easy conversation. That's a reasonable way to hire for the relationship accounts. It's a terrible way to hire someone whose real job is solving technical problems for hours at a time.
Then came onboarding. Same classroom for everyone. Same high-energy, cheerleader-style training, no matter the role.
Picture the quiet, detail-minded new hire who was perfect for the technical accounts, sitting through days of cheering and group games. Now picture the outgoing new hire for the relationship accounts, alone with a thick manual. Both start wondering whether they made a mistake.
They were not hiring bad people. They were leading everyone through a poorly planned experience and reading the results as a character flaw.
What we changed
We started from the roles, not the resumes. We built a behavioral model for each type of position and separate interview guides for each one. Then we changed how the training was delivered, so it matched how each person learns.
Turnover dropped immediately. People were happier, and they stayed. Client satisfaction scores rose sharply, and their clients started sending referrals.
Nothing about the labor market changed. Nothing about the applicant pool changed. We stopped running people through an experience that was never designed for them. You can read the full story here.
Interviews that find fit, not polish
Onboarding can only work with the people your interviews let in, so start there.
- One interview guide per role. If two roles demand different things, they need different questions. Ask about the real work, not about the candidate's favorite strengths.
- Interviewers who know their own drives. Without realizing it, people reward candidates who interview the way they do. Once your interviewers know how they're wired, they can catch themselves doing it.
- Polish is not fit. A great interview tells you someone interviews well. It doesn't tell you whether they'll want to do this work at 4 o'clock on a Thursday.
Hire for the role. Then onboard the person you hired.
Onboarding built for how people learn
You don't need a different program for every person. You need to notice that people learn differently, and give them a few options. The four drives are a good place to start.
Agree on the outcomes and what good work looks like.
Know what support and communication they need.
Make the expectations clear before measuring performance.
Ask what is working and adjust together.
- Dominance. People with more of it want ownership early, so give them a real problem in week one. People with less want to know who will support them, so give them a buddy.
- Extraversion. People with more of it learn with people. Pair them up, let them shadow and let them talk it through. People with less learn on their own first. Give them time with the material before the group session, and written follow-up after it.
- Patience. People with less of it want quick wins and variety, so give them early goals they can hit. People with more of it want to know what's coming, so give them the schedule for the first 30 days on day one.
- Formality. People with more of it want the manual, the checklist and the reasons, so hand them the details. People with less want the headline and to learn by doing, so show them what good looks like and let them try.
Placement matters too. Put a very social person in a role with almost no interaction, and they won't last. If they applied for it anyway, they probably wanted the paycheck, not the work.
The 4 o'clock test
Walk past your team at 4 o'clock this week. Who has everything packed and is waiting for the clock? Who is still in the middle of something because they want to finish it?
The right people stay because they want to, not because they have to.
The first group isn't lazy, and they aren't bad people. Most of the time they're in a role, or an experience, that doesn't fit how they're wired. That can be fixed.
Five questions for your onboarding
Ask these this month:
- Does every new hire get the same training, no matter the role?
- Who designed your onboarding, and how are they wired? Training tends to look like the person who built it.
- Do your interviewers reward people who interview the way they do?
- When do people leave, and what were they doing that week?
- Does each new hire's manager know how that person learns best?
If those questions make you uncomfortable, good. That's where the work is.
When you're ready to define what each role really demands before you hire, start with Job Targeting. To see the whole approach, from the first conversation to a team that stays, read how it works.
Blame is easy. Leadership is not.
Questions readers ask
Hire for what each role really demands, onboard people the way they learn, and lead each person the way they need to be led. Most turnover is decided long before someone resigns.
Often because the role, the interview or the onboarding was built for a different kind of person. When someone has to fight their natural drives from day one, they start wondering whether they made a mistake.
A clear schedule for the first 30 days, a mix of learning with people and learning alone, early goals people can hit, the details for those who want them, and a manager who knows how each new hire learns best.
No. The costly kind is losing people who fit the role and would have stayed if their first weeks had been built for them.
Mike McCloskey
Founder and CEO of HumanLytics Team. Helping leaders turn people data into a people plan.
More about Mike →